Booth, G. D. & Miller, H. R. (1974). Effectiveness of monochrome and color presentations in facilitating affective learning. AV Communication Review, 22, 409-422.

(Conclusions.) The assumption that prior years of monochrome viewing would work to the detriment of color viewing in the matter of learning affect (negative transfer) is unsupported.
Results of the survey of the pupil home viewing mode in three suburban areas indicated a ratio of color to monochrome that was reasonably close to one-to-one.
Pupils in grade two appeared to show, in a greater number of instances, more internalization (involvement) toward positive attitudes when viewing the monochrome presentation. Pupils in grades two and six have apparently employed more processes (energy) leading to ideation, or imagination, when viewing monochrome than their counterparts who had viewed color.
Generally, pupils in grade four appear to be in a transitional state and the statistical and observational data resulting from this investigation support no definite conclusions or explanations.
The color variable may be a positive factor in promoting levels of valuing in grade six.
Inspection of the implications for the narrative vs. non-narrative and its ability to raise items of statistical significance become purely speculative in that any consideration is also dependent upon question or item formulation.
There appears to be a steady decay in attitude toward desirable educational activities (as viewed by the jury) from the positive position in grade two, to the negative in grade six.
Mode of presentation seems to have some effect on opinion, but any specificity of interaction between the two falls into the area of conjecture at the present time.
It must be remembered that the black-and-white images were displayed on a television set and that the color images were front projected onto a screen. Students may have responded to the "surround" of the images, as well as to the images themselves. In other words, the students viewing the balck-and-white presentations were watching television programs, while the students viewing the color presentations were watching a motion picture (or movie). Are there any associations here that may have influenced their responses? A similar experiment by Link (1961) does not address this question either. Certainly any replication of this study should take this problem into consideration.